Data audit
| file | SHA-256 | rows x cols | missing | duplicates |
|---|---|---|---|---|
| data/endowments.csv | fc130a2f54c750a5834ec7314060d91f7984769c7c006ad04b98582a016ba74b | 2 x 3 (agent, good1, good2) | 0 | 0 |
| data/parameters.csv | 717807e75337fbf4aff8e4d74e0fc0b83b799f78fa689172be0d501cb3e43983 | 4 x 3 (parameter, value, description) | 0 | 0 |
Hashes match bundle.json. Units: units of each good (theoretical). No dates.
- Endowments: A=(8,2), B=(4,8); all strictly positive.
- alpha_A=0.4, alpha_B=0.6, both in (0,1) -> interior Cobb-Douglas demands.
- Anomaly: aggregate_endowment_good1 recorded = 14.0 but sum of individual endowments = 8+4 = 12.0. aggregate_endowment_good2 recorded = 10.0 = 2+8 (consistent).
- Decision: the per-agent endowments are designated primary data in TASK.md; the summary sheet is secondary. Market clearing uses 12.0 (the sum); the inconsistency is reported in data_issues. Clearing against 14 while agents' wealth is computed from (8,4) would violate Walras' law, so 14 cannot be internally consistent.
- No availability-rule violations (only data/ used).